Обсуждение: pgsql: In the stats test, delay for the stats collector to catch up

Log Message:
-----------
In the stats test, delay for the stats collector to catch up using a
function that actually sleeps, instead of busy-waiting.  Perhaps this
will resolve some of the intermittent stats failures we keep seeing.

Modified Files:
--------------
    pgsql/src/test/regress/expected:
        stats.out (r1.4 -> r1.5)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/expected/stats.out.diff?r1=1.4&r2=1.5)
    pgsql/src/test/regress/input:
        create_function_1.source (r1.14 -> r1.15)

(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/input/create_function_1.source.diff?r1=1.14&r2=1.15)
    pgsql/src/test/regress/output:
        create_function_1.source (r1.25 -> r1.26)

(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/output/create_function_1.source.diff?r1=1.25&r2=1.26)
    pgsql/src/test/regress:
        regress.c (r1.62 -> r1.63)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/regress.c.diff?r1=1.62&r2=1.63)
    pgsql/src/test/regress/sql:
        stats.sql (r1.2 -> r1.3)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/test/regress/sql/stats.sql.diff?r1=1.2&r2=1.3)

Tom Lane wrote:
> Log Message:
> -----------
> In the stats test, delay for the stats collector to catch up using a
> function that actually sleeps, instead of busy-waiting.  Perhaps this
> will resolve some of the intermittent stats failures we keep seeing.

Good.  I was seeing these random stat failures too.

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

On Sat, Jul 23, 2005 at 11:18:57AM -0300, Tom Lane wrote:
> In the stats test, delay for the stats collector to catch up using a
> function that actually sleeps, instead of busy-waiting.  Perhaps this
> will resolve some of the intermittent stats failures we keep seeing.

Any chance of exposing a sleep function to the user?  It's admittedly
of limited utility and it can usually be done on the client side,
but I've occasionally wanted one for testing with simple SQL scripts.
In such cases I end up writing a one-liner with PL/Perl, PL/Tcl, etc.

--
Michael Fuhr
http://www.fuhr.org/~mfuhr/